Molecular biomarkers for seabird age estimation : implications for ecological monitoring

Seabirds are widely used as an indicator species to track important changes in ecosystems. The key data required for monitoring include (i) population trends (ii) status, including demographic properties such as age structure and reproductive performance. Most seabirds have no outward or easily identifiable marks to age individuals and there are currently no robust molecular methods for age estimation in birds. Instead, individuals must be marked by leg rings as chicks to establish known-age populations, which is a labour intensive and expensive process. Long-term monitoring of some populations of birds has produced only small numbers of individuals for which there is a known age, highlighting the urgent need to identify alternative aging techniques. In this thesis, I describe the development and testing of age-related differentially methylated positions (aDMPs) as a minimally invasive method for the estimation of chronological age in the Short-tailed shearwater (Ardenna tenuirostris). Recent studies have revealed novel examples of DNA methylation (DNAm) age association in several mammalian species, indicating the potential for developing DNAm age biomarkers for a broad range of wild animals. Emerging technologies for measuring epigenetic signals have also enhanced our ability to study age-related DNAm changes. I assessed DNAm in several shearwater genes that have shown age-related DNAm changes in mammals. In birds ranging in age from chicks (months old) to 21 years, bisulphite treated blood and feather DNA was sequenced. The sequences allowed DNAm analysis across 67 CpG sites in 13 target gene regions. Despite the identification of some weakly correlated aDMPs, the majority had no clear association with age and statistical analysis using a penalised lasso approach did not produce an accurate ageing model. Our data indicated that some age-related signatures identified in orthologous mammalian genes are not conserved in the shearwater and that an alternative technique would be required to progress in this area. The alternative approach that I tried was digital restriction enzyme analysis of methylation (DREAM). I quantified DNAm in whole blood samples from a total of 71 known-age shearwater using DREAM. This method measures DNAm levels at thousands of CpG dinucleotides throughout the genome. We identified seven CpG sites with DNAm levels that correlated with age. A model based on these relationships estimated age within a mean of 2.8 years of known age, based on validation estimates from models created by repeated sampling of training and validation data subsets. Longitudinal observation of individuals re-sampled over 1 or 2 years generally showed an increase in estimated age (6/7 cases). These seven aDMPs were then selected for use in a more cost-effective, targeted approach using a next generation sequencing assay. I validated and then generated chronological age estimates in a large set of known, minimum and unknown age individuals sourced from our study population. Using targeted NGS we confirmed age-correlation in three of the aDMPs identified using the DREAM method. We also identified an additional 14 aDMPs that were within close proximity to two of the original markers described above. Using four aDMPs to create a shearwater epigenetic age assay we trained a model on 109 known-age blood samples. DNAm age estimates of minimum-age individuals (N = 55) showed that 39/55 (71%) had estimates above their known minimum-age. A further 23 unbanded individuals from a nearby island were used to create a small population age structure. This age structure, although small, showed similar age-class assignment to the known structure determined from long-term ringing. For the first time, we have shown that epigenetic changes with age can be detected in a wild bird. The ability to differentiate age classes e.g. chicks, young breeders and birds of middle and old ages, of previously unbanded seabirds will be a useful tool for scientists and managers. Estimated age data could be informative for monitoring of post-pest eradication around island-breeding seabird populations or for determining impacts of longline fishery interactions. The molecular techniques developed for age estimation in this thesis also provide fundamental knowledge on seabird and mammal gene and CpG conservation, DNAm quantification and analysis in non-model animals.